A woman writing notes by hand in a small notebook with a pen

AI and Web Skills: Why Building by Hand Still Matters

Picture a small web team at the end of another busy week. They have shipped more pages, more features and more fixes than ever before, largely thanks to AI coding assistants. Yet nobody feels more confident in their craft, and quite a few feel worn out. That picture is becoming familiar right across the web industry.

A growing number of experienced voices in front-end development are now asking a simple question: if AI has changed how we work this quickly, what should designers, developers and the people who manage them do to stay sharp? The answer turns out to be less about the tools and more about people.

A man resting his head on his arms on a desk beside a closed laptop, looking exhausted

Speed has gone up, satisfaction has gone down

Over the last couple of years, big tech firms and AI start-ups have pushed large language models into almost every corner of digital work. Plenty of businesses responded by telling staff to use AI for as much as possible. Some of those top-down mandates have since softened, as companies found the technology expensive to run and the promised returns slower to arrive than expected.

What hasn’t gone away is the pressure for volume. Teams are often expected to spend the hours AI supposedly saves on producing even more, whether or not that extra output is needed. Developers describe days spent checking and patching machine-written code rather than building things themselves, and many say the job has lost much of its enjoyment.

Go Ad Free

There is a wider cost too. The web grew quickly because people shared what they knew freely, through blog posts, open-source projects, tutorials and community conferences. Independent educators and publishers who keep that culture going report shrinking audiences and income as more people turn to chatbots for answers.

Myths and facts about AI and your skills

  • Myth: If AI can write the code, there’s no point learning it properly.
    Fact: What makes a good developer or designer has never been typing speed. It’s experience, judgement, an eye for nuance and creativity – exactly the qualities needed to spot when generated code is wrong.
  • Myth: Time saved by AI should go straight back into more output.
    Fact: Filling every saved hour with more tasks is a quick route to burnout. Early academic research has already linked heavy, daily reliance on chat-based AI tools with fatigue and weaker critical thinking.
  • Myth: AI will keep improving dramatically, so investing in people is a waste.
    Fact: Even many enthusiasts accept that the gains between new models are levelling off. People who understand the fundamentals remain the safest investment.
  • Myth: Today’s low prices for AI tools are here to stay.
    Fact: Current pricing is widely believed to be propped up by investor money. If costs rise, workflows built entirely around cheap AI output could suddenly become expensive, particularly when its mistakes need fixing.
A small team on laptops watching a colleague explain ideas on a whiteboard covered in sticky notes

Five ways to keep human skills sharp

The practical advice is aimed at two groups: the individuals doing the work, and the founders and managers who decide how that work gets done. Both have a part to play.

  1. Protect time for learning. The web industry was once excellent at upskilling its people. Give staff paid time and a budget for courses, reading and conferences, rather than treating every spare hour as spare capacity.
  2. Review any AI mandates. Ask whether pushing for maximum output still makes sense once the true costs are counted. Bring the people affected into that conversation and listen to how the policy feels from their side.
  3. Build something by hand. Repetition is how knowledge sticks. Writing HTML and CSS yourself, at a slower pace, gives your brain room to think, and that shows in the quality of the finished work. A small side project made without any AI help is a good place to start.
  4. Make it part of the working week. Managers can let people do this kind of practice on the clock. It costs a little time and pays back in confidence and morale.
  5. Support the people who teach. Courses, memberships and independent learning resources keep the sharing culture alive. Paying for good training is far cheaper than rebuilding lost skills later.

A real-world example

Imagine a local shop asks for a new booking page. One approach is to prompt an AI assistant, paste in whatever comes back and move on. The page might look fine, but if nobody truly understands the code, every future change becomes a gamble, and accessibility problems or broken layouts on certain phones can slip through unnoticed.

Now imagine a developer who knows their HTML, CSS and accessibility basics inside out. They may still use AI to speed up a repetitive job, but they can read the output critically, catch the errors and explain their choices to the client. Their page takes a little longer to build, yet it is easier to maintain, cheaper to fix and far less likely to let the business down.

For small businesses, that difference matters. When you hire someone to build or look after your website, it’s worth asking how they use AI and, more importantly, how they check its work. The tools will keep changing; a skilled person who understands what’s under the bonnet is still what keeps a website running well.

Go Ad Free

Create a free account, or log in.

Gain access to read this article, join developers who build better, faster, smarter solutions

Yes! I would like to receive new content and updates.

Scroll to Top